defgate: No validation whether the matrix is unitary

Hi,

This is not a feature request, I’m just trying to wrap my head around the infrastructure - if this is not the right place to ask, please let me know where I can ask learning questions as a total newbie to quantum computing and the Rigetti stack.

I was wondering what is the reason behind that I can create a quantum gate based on a non-unitary matrix? Is this something I’d have to be careful about in the future as well, or Forest is planned to have a feature to validate against non-unitary operators? Also, it’s one thing that it can run on the QVM - but how would this operate on a real quantum processor?

>>> import pyquil.quil as pq 
>>> import numpy as np
>>> import pyquil.forest as forest
>>> qvm = forest.Connection()
>>> non_unitary_mx =np.array(([1.0, 2.0], [3.0, 4.0]))
>>> bad_prog = pq.Program().defgate("NON-UNITARY-BAD-GATE", non_unitary_mx)
>>> bad_prog.inst(("NON-UNITARY-BAD-GATE",0))
<pyquil.quil.Program object at 0x1096e92d0>
>>> bad_prog.inst(("NON-UNITARY-BAD-GATE",1))
<pyquil.quil.Program object at 0x1096e92d0>
>>> print bad_prog
DEFGATE NON-UNITARY-BAD-GATE:
    1.0, 2.0
    3.0, 4.0

NON-UNITARY-BAD-GATE 0
NON-UNITARY-BAD-GATE 1

>>> print qvm.wavefunction(bad_prog)
(array([ 1.+0.j,  3.+0.j,  3.+0.j,  9.+0.j]), [])

[stylewarning]

I thought I might note what the Quil spec says about this.

The gate is declared using the DEFGATE directive followed by comma-separated lists of matrix entries indented by exactly four spaces. Matrices that are not unitary (up to noise or precision) have undefined [7] execution semantics.

[7] Software processing Quil is encouraged to warn or error on such matrices.

Accordingly, an implementer of Quil can define how he or she pleases what happens when someone defines a non-unitary gate.

[willzeng]

Closed by #21